A123 20Ah prismatic cells got expanded - What can be done?

Joined
Mar 29, 2016
Messages
383
Hi,
I use those cells on all my E-bikes.
I wrap all of them into a battery pack using several layers of gopher tape, which proved very well for applying the required pressure upon the cells. (I apply pressure by hand when I wrap them, as they require to be pressurized a bit)
Yesterday, I discovered that half of the cells inside the 10S pack on the touring bike got expanded.
So now I several unsolved dilemmas which I couldn't find solution for here by doing a search:

1. I assume that the shocks and vibrations on that bike (700x45 tires at 60PSI with no suspension) caused that? (The other packs on the other bikes which are sprung looks fine)
2. I know that these cells will vent in case the gas pressure builts up - so I wonder if any better pressurization of the cells would have helped (for example if they were inside a rigid metal case) since that gas has to go somewhere...
3. Will a rigid metal enclosure prevent that gas-out in the first place, in-spite of those vibrations due to lack of suspensions, or do these cells not recommended in such scenario in the first place?
4. Can the expanded cells be fixed some how by forcing the gas to be recombined with the electrolyte? (Each cell expanded just by a bit, but together as a 5 - that protrusion looks much more significant).
5. If not - Are they still usable? And what should be the order of the capacity loss?
6. I am curious why exactly half of them expanded (the left side of the battery, inside the frame triangle), although I guess this would remain a mystery...
 
With pouch cells of any persuasion, their biggest threat is physical trauma. Vibration, fretting, rubbing and puncture are real risks which need to be mitigated against. Unfortunately gaffa tape is not as good as a hard enclosure.
The gassing happens when a cell is heated excessively - the electrolyte is typically ethylene carbonate or propylene carbonate. These solvents may release CO2 when heated above a certain point. So the good news is that the gas is harmless, but the bad news is that gaseous pockets between the layers of active material simply increases their internal resistance and shortens their life. You will probably find that they either self-discharge more than usual or you've just got a reduction in capacity.
You may find they will continue to run reasonably well, but if I were you I'd put them inside a rigid enclosure and fix them inside with an adhesive to prevent rubbing or fretting. If you use a metal for this enclosure, make sure you line the inside with something like Kapton tape so that it can't short on the cells.
 
and there are a number of threads about a123 pouch cells, including termination and packaging, which might help you figure out a way to keep this from happening to any other cells.

there are specific pressures that need to be applied to them, evenly across their faces (not the edges), which tape won't do by itself (or possibly at all). posts by wb9k have good info about this kind of stuff.

some of the threads are in this list
https://endless-sphere.com/forums/search.php?keywords=a123&terms=all&author=&sc=1&sf=titleonly&sr=topics&sk=t&sd=d&st=0&ch=300&t=0&submit=Search
though not everything there is relevant
 
amberwolf said:
and there are a number of threads about a123 pouch cells, including termination and packaging, which might help you figure out a way to keep this from happening to any other cells.

there are specific pressures that need to be applied to them, evenly across their faces (not the edges), which tape won't do by itself (or possibly at all). posts by wb9k have good info about this kind of stuff.

some of the threads are in this list
https://endless-sphere.com/forums/search.php?keywords=a123&terms=all&author=&sc=1&sf=titleonly&sr=topics&sk=t&sd=d&st=0&ch=300&t=0&submit=Search
though not everything there is relevant

Thanks for replying. I was already reading about them here on ES. I wanted to ask if anything can be done in the current situation case, eg - if it's possible to recover them.
It only happened on this bike, and only on the battery which sits on the unsprung frame. The handle-bar battery which is sprung is fine, as well as all my other A123 20Ah batteries on all the other E-bikes I have. (as they are sprung mass)
I tend to believe that even with the tightest packing, and with the right pressures, these cells were just not meant to withstand the unsprung shocks and vibrations of a road bike, so they would eventually fail - and if they can't expand to contain the gas, they will just vent it out. (which will be obvious with time only when significant capacity is lost). Am I right about this?
 
jonescg said:
With pouch cells of any persuasion, their biggest threat is physical trauma. Vibration, fretting, rubbing and puncture are real risks which need to be mitigated against. Unfortunately gaffa tape is not as good as a hard enclosure.
The gassing happens when a cell is heated excessively - the electrolyte is typically ethylene carbonate or propylene carbonate. These solvents may release CO2 when heated above a certain point. So the good news is that the gas is harmless, but the bad news is that gaseous pockets between the layers of active material simply increases their internal resistance and shortens their life. You will probably find that they either self-discharge more than usual or you've just got a reduction in capacity.
You may find they will continue to run reasonably well, but if I were you I'd put them inside a rigid enclosure and fix them inside with an adhesive to prevent rubbing or fretting. If you use a metal for this enclosure, make sure you line the inside with something like Kapton tape so that it can't short on the cells.

There was no heat-up at all. Zero. On my hungriest E-bike, which can draw and sustain current in the order of 90A (5C for 20Ah) the cells get barely warm thanks to their low IR. I am pretty sure now it's due to the vibrations of a road-bike, which no container can withhold.
 
In that case yeah most likely vibration and fretting.
The best you can do is fix it firmly to the inside of the enclosure, and fix the enclosure firmly to the chassis of the vehicle. Last thing you want is a cell moving independently of the chassis, even if there's a layer of foam/rubber there to dampen the movement.
 
I've read about making a small puncture with a pin and squeezing the gas out, then seal the hole with epoxy. I can't recommend this, but most accounts are that this works. Having too much gas in the pouch will make it hard to get proper compression on the plates.
 
jonescg said:
In that case yeah most likely vibration and fretting.
The best you can do is fix it firmly to the inside of the enclosure, and fix the enclosure firmly to the chassis of the vehicle. Last thing you want is a cell moving independently of the chassis, even if there's a layer of foam/rubber there to dampen the movement.

Do you think that even with proper compression and immobilization, the cells would not degrade due to the excessive vehicle shocks?
I mean - yes, they would be constrained, but they would still "feel" everything as they would be rigid with the bike-frame. (Again, road bike frame that vibrates a lot)
I am assuming that the gas out will still take place, except that with proper compression pressure it will have nowhere to go, so eventually the cell would vent it slowly, causing slow but gradual capacity loss. OR, you say that with proper compression+immobilization the gassing-out would not take place even with excessive vibrations, and so the cell could reach it's "life expectancy" ?
 
fechter said:
I've read about making a small puncture with a pin and squeezing the gas out, then seal the hole with epoxy. I can't recommend this, but most accounts are that this works. Having too much gas in the pouch will make it hard to get proper compression on the plates.

I heard from Russian speaking friend who red it in a Russian-speaking forum about freezing such cells and then immediately compressing them when they are cold so the gas takes much less volume, allowing easier compression. Why I mention "Russian"? Because such discussions are nearly impossible to find on Google, and so he found it on some remote Russian speaking forum, which as a non-Russian speaker I can't read... :?

But still, once some electrolyte has decomposed into gas, some capacity is lost so I can't series connect such cells with the remaining healthy cells right? Or, perhaps by freezing them and then compressing I could force the other-way-round chemical reaction and restore the capacity+IR? (As happens in Nickel based battery)
 
As Fechter said, they can be pricked and de-gassed but it's a risky process. You're just as likely to introduce water and air which won't help. I think AussieJester did a video of himself doing this for a puffed LiPo. Compression is critical with pouch cells. Not multi-gigapascal compression, but snug at a bare minimum. Not sure about restoring the low IR - has anyone ever tested that?

As long as the cells are not moving relative to whatever they are in contact with, they will be fine. Shock loads are just really rapid accelerations and decelerations. The problem is when you run into something, or chafe across something. A bit like going hiking without socks - eventually your skin will rub and blister. The socks are there to prevent movement of your foot relative to the inside of your boot.

I view pouch cells as being like unpeeled bananas. They are fine as is, provided nothing touches them. But once inside their purpose-built enclosure (peel) they are very rugged and can survive a bit of rough and tumble.
 
thunderstorm80 said:
fechter said:
I've read about making a small puncture with a pin and squeezing the gas out, then seal the hole with epoxy. I can't recommend this, but most accounts are that this works. Having too much gas in the pouch will make it hard to get proper compression on the plates.

I heard from Russian speaking friend who red it in a Russian-speaking forum about freezing such cells and then immediately compressing them when they are cold so the gas takes much less volume, allowing easier compression. Why I mention "Russian"? Because such discussions are nearly impossible to find on Google, and so he found it on some remote Russian speaking forum, which as a non-Russian speaker I can't read... :?

But still, once some electrolyte has decomposed into gas, some capacity is lost so I can't series connect such cells with the remaining healthy cells right? Or, perhaps by freezing them and then compressing I could force the other-way-round chemical reaction and restore the capacity+IR? (As happens in Nickel based battery)

Ideally use a dry-box or something purged of O2 and water vapor both if possible. The glues to re-seal the hole can be rough to ensure you got it and it lasts against the carbonate ester electrolyte. I've never tried to do it with A123 pouches or electrolyte, so I don't have a suggestion for a glue to use. I've seen success with shu-glue and some high temp hot melt glues on slightly roughed up surfaces wiped with acetone just before gluing.

It's not the kind of thing to recommend to someone who isn't comfortable with the idea of the battery randomly burning before, after, or during, and if the cell has damage causing it to make a bunch of gas, just venting the gas out is a temporary band-aid, because it's going to fill up again fairly soon.
 
liveforphysics said:
thunderstorm80 said:
fechter said:
I've read about making a small puncture with a pin and squeezing the gas out, then seal the hole with epoxy. I can't recommend this, but most accounts are that this works. Having too much gas in the pouch will make it hard to get proper compression on the plates.

I heard from Russian speaking friend who red it in a Russian-speaking forum about freezing such cells and then immediately compressing them when they are cold so the gas takes much less volume, allowing easier compression. Why I mention "Russian"? Because such discussions are nearly impossible to find on Google, and so he found it on some remote Russian speaking forum, which as a non-Russian speaker I can't read... :?

But still, once some electrolyte has decomposed into gas, some capacity is lost so I can't series connect such cells with the remaining healthy cells right? Or, perhaps by freezing them and then compressing I could force the other-way-round chemical reaction and restore the capacity+IR? (As happens in Nickel based battery)

Ideally use a dry-box or something purged of O2 and water vapor both if possible. The glues to re-seal the hole can be rough to ensure you got it and it lasts against the carbonate ester electrolyte. I've never tried to do it with A123 pouches or electrolyte, so I don't have a suggestion for a glue to use. I've seen success with shu-glue and some high temp hot melt glues on slightly roughed up surfaces wiped with acetone just before gluing.

It's not the kind of thing to recommend to someone who isn't comfortable with the idea of the battery randomly burning before, after, or during, and if the cell has damage causing it to make a bunch of gas, just venting the gas out is a temporary band-aid, because it's going to fill up again fairly soon.

After a full day in the freezer at -20oC degrees, the cells which were a bit inflated are now completely flat. They are still a bit soft to the touch and not like a hard-surfaced healthy cells, but it seems it's totally possible to compress them in this way and constrain the gas within. I also think in this way the gas would add it's own pressure once it's warmed back up, or, hopefully, it will be dissolved back into the electrolyte. :roll:
I will try compressing them with two thick aluminium plates and metal zip ties, and see how it goes...
 
Sadly, it won't dissolve back.
 
liveforphysics said:
Sadly, it won't dissolve back.

So I just dump those cells, or, use 4 of them as an emergency car starter.
I will rebuild the pack with spare good cells I have, and will press it with two aluminium plates.
I understand that with proper pressure, even the massive shocks and vibrations of road-bike frame wouldn't degrade them? (They would be pressed against each other (about 10PSI), but the whole pack will still vibrate and receive shocks as it would be rigid with the frame.
What do you think ?
 
I have two 12s A123 20ah packs on my bike for 4.5 years and 1,088 cycles. I use at 40 to 90amp now. I get 17ah at that level. Bulk charge to 3.5v and rarely need balance. I had replaced one in the beginning as it puffed and got soft.
I built the pack with plastic totes and fiber duct tape. The one on the back rack has a thin plywood and rubber underneath. I ride the rattlesnake/horse trails and low tide beach. 40mph on a gaint hard tail on the street. When I get in trouble I jump off the bike and let it go (3) times. The center 12s just has plastic sides roped in with Velcro and a plastic gutter underneath. I love these cells even though they are so heavy. Would like a 16s in a trianlge if I could get them cheap.
I don't like playing with compromised cells as I don't have a safe place like a BBQ for safe sleeping.
Good luck
P.S. love those A123's
 
A123 20ah as 4s for a car battery sounds good. Like to know how it works out and what size car. 3.5 × 4 = 14 volt high - 2.9 × 4 = 11.4 volt low . A car doesn't have lvc.
 
999zip999 said:
A123 20ah as 4s for a car battery sounds good. Like to know how it works out and what size car. 3.5 × 4 = 14 volt high - 2.9 × 4 = 11.4 volt low . A car doesn't have lvc.

I mean to use them as an auxiliary 12V battery for emergency - for example to jumpstart a car with a drained lead battery or for a portable power source for appliances that can run on 12V in case of a blackout. (Charge your phone using a cheap carjack 12V->5V)
With 12V*20Ah you can power many smartphones and some laptops when you go out camping...

I also thought about using it as THE car starter battery but it has it's caveats: No LVC as you said, high-temperature working environment, no overcharge protection and a charging "algorithm" which is designed for Lead, but the biggest issue of them all is that a modern parked car draws a significant current from the huge 720Wh lead battery (immobilizer, alarm, etc..), and they will draw a tiny 12*20=240Wh in several days if unused or within 2 hours if you forgat your headlights on. Having a 720Wh A123 battery would still weight MUCH less than lead, but that requires 12 spare cells. Not 4.
 
999zip999 said:
I have two 12s A123 20ah packs on my bike for 4.5 years and 1,088 cycles. I use at 40 to 90amp now. I get 17ah at that level. Bulk charge to 3.5v and rarely need balance. I had replaced one in the beginning as it puffed and got soft.
I built the pack with plastic totes and fiber duct tape. The one on the back rack has a thin plywood and rubber underneath. I ride the rattlesnake/horse trails and low tide beach. 40mph on a gaint hard tail on the street. When I get in trouble I jump off the bike and let it go (3) times. The center 12s just has plastic sides roped in with Velcro and a plastic gutter underneath. I love these cells even though they are so heavy. Would like a 16s in a trianlge if I could get them cheap.
I don't like playing with compromised cells as I don't have a safe place like a BBQ for safe sleeping.
Good luck
P.S. love those A123's

You also don't have BMS, right?
I also charge them to 3.5V@cell, and top-out all the cells to 3.6V once in a while using a small Turnigy balance charger.
BUT, I realized that with aging the cells lose the capacity differently, and this means that some cells exhaust before others as the time goes by. If you never fully discharge it's not a problem, but a 100% discharge could result in some cells going much below 2V.
Therefore, I thought to find a voltage monitor which I can plug in to the external 6S balance ports I have installed on each pack, and so I can have a visual output (for example using 6 LEDs that can light-up/change color when a certain cell drops below 2~2.5V), to know when a certain cell in a group has exhausted so I know when to stop draining the pack.
Do you know of such product?
Yes, I could build something like this by myself, but I would prefer a turnkey product for reliability. And it should be something which is powered ONLY from the balance port. (And since all balance ports would be drained this way I don't see any balance drifting issues)

I also understood that according to the spec the 2V cut-off is a recommended cut-off, and they only strictly advice not to go below 0.5V.
Of course no one would design a system so it would drain the cells below 2V on each cycle (I personally set my LVC to 2.5V@cell), but I understand that if for example due to balance issues/capacity difference some cells drained to 1V-1.5V or so, and if I follow with immediate charge - then there won't be any noticeable damage, right?
 
Nio bms bullshit. My battery is dead at 2.8v and 3.45v is great not much above. Yes the time I have trouble is at lvc. By the time I get off my bike and check there at 2.8v I imagine they hit 2.7. - 2.65v. They don't like being at the bottom in that's when I could feel they lose a little capacity. 1,098 cycles just looked. I bluck charge each 12s @ 16amp with a hp 600 power supply and a meanwell set to 29volt in series times two on one power strip. Just have four 6s sense wires that have been replaced 4 times over the years, because I plug it in and check cell voltage every charge or at least half of one 12 oz pack because everyone's at 21.124v the other one's got to be close. I check with cheap 6s monitors. You could use a 4s alarm and set lvc. Or one hour for 18ah of 24s.A123 20ah cells from Victpower cheap ass company. Beware and feel lucky to gamble with that company.
 
Plus where did you get these cells what kind of tabs they have on them to someone add on some tabs and cut them out of a pack. New used and how old are they mine were probably made in 2010 or something I got him 5 years later or whatever that makes
 
999zip999 said:
Plus where did you get these cells what kind of tabs they have on them to someone add on some tabs and cut them out of a pack. New used and how old are they mine were probably made in 2010 or something I got him 5 years later or whatever that makes

I got them second hand a year a go from a friend who bought them from China. They are from a company which makes copy, by license from A123, not the original A123. They have around 17.5Ah when discharging them at 40A-70A.
 
I still didn't get an answer for the most important question:
The cells got inflated obviously because of the vibrations and the shocks of the road bike frame + their lack of pressure container.
I am wondering, if those cells were meant to work at such unsprung environment even if I pack them with rigid enclosure and with the right pressures? I mean, will such environment will shorten their life significantly ? (And so it's better to pick an 18650 battery for that application)
Please answer this :D
 
As far as I know, there should be no problem with vibration as long as the cells are properly constrained. Zero motorcycles use prismatic cells these days (although Li-ion chemistry). This is a high vibration environment also but they hold up fine. Their packs are now fully potted.
 
fechter said:
As far as I know, there should be no problem with vibration as long as the cells are properly constrained. Zero motorcycles use prismatic cells these days (although Li-ion chemistry). This is a high vibration environment also but they hold up fine. Their packs are now fully potted.

The cells inside the zero motorcycle are part of the sprung mass, while the road bicycle isn't.
I use those cells on all my E-bikes, and the puffing took place only in the road bike. (All the other E-bikes have either full suspension, or 10-15PSI fat-bike tires)
I used only duct-tape to constrain them, and this is why I wonder if the problem is the excessive vibration of the unsprung mass, and will a rigid enclosure+proper pressure would protect them in such environment. I didn't find any documentation about this.
 
Like I said I bounce my bike for 4.5 year old pack put together with plastic totes and duct tape I've dumped my bike 3 times I beat the s*** out of it on the trail. I believe you have contaminants inside the cell and or impurities. Maybe overcharge over discharge. But vibration on a road bike I don't think so.
I personally would not puncture and reseal the pouch as I do not have a place to safely store and charge them.
 
Back
Top